Today I read a paper about bee population dynamics published in Ecology (Franzen and Nilsson 2013). Given the current concern about bee declines (more on that in a few weeks) one can assume we (scientists) understand the basic dynamics of bee populations, or at least we have an idea of their life histories. Well, the paper monitored one metapopulation of one species during 9 years and found that fluctuations on the number of nests among years are huge (more than one order of magnitude). Why? We don’t know and It is not correlated with floral resources or climate. Some speculations include source-sink dynamics, a prolonged diapause or bet hedging strategies to avoid natural enemies. We know nothing. And you may ask, why is this published in Ecology? Well, because I think is a good paper that at least shows some data. That means that given the knowledge we currently have, this tiny bit of information advances our understanding.
Really basic research is not sexy but can we (and I am the first guilty) understand a pollinator crisis if we don’t know if it is predation or it is competition what is driving bee fitness. Or can we understand the actual structure of plant-pollinator networks, which are characterised by an incredible turn over among years, without knowing if bed hedging strategies are the norm or the exception (Danforth 1999). Can we assess the effect of landscape configuration on bee populations without the basic natural history information like eggs per female, or growth rates?
Franzen M. & Nilsson S.G. (2013). High population variability and source-sink dynamics in a solitary bee species, Ecology, 130204095918002. DOI: 10.1890/11-2260.1
Bartomeus lab 
Great post, well said! As a friend once said, mixing his metaphors, in ecology we have only scratched the surface of the tip of the iceberg.
We see this in butterfly populations as well. Typically due to a combination of weather, genetics (allowing for emigration), and parasitoid populations.